Abstract Background To perform electrical cardioversion of atrial fibrillation or atrial flutter with a duration of more than 48 hours, it is recommended to initiate anticoagulation therapy for at least three weeks beforehand. In cases where cardioversion is necessary earlier, it is suggested to precede the procedure with a transoesophageal echocardiography to rule out the presence of a thrombus in the left atrium (referred to as TOE-guided cardioversion). This strategy is outlined in current ESC guidelines for atrial fibrillation. However due to limited evidence, no specific level of recommendation for its use is provided. Purpose This study aimed to characterize patients treated by TOE-guided cardioversion in two hospitals and to conduct a safety analysis of the procedure. The 30-day incidence of thromboembolic events was chosen as the main safety objective, the 30-day mortality as the secondary. Method This was a retrospective, two center, cohort register study where information was obtained from patients medical records. All patients who underwent TOE-guided cardioversion between 2012-2020 in the first institute, and between 2015-2020 in the second were included. From the medical records all the patient characteristics and the clinical outcomes were identified. Result A total of 259 patients, with TOE-guided cardioversions, arrhythmia duration over 48 hours and without adequate anticoagulation therapy were included. Patient characteristics are shown in table 1. It was equal incidence of a 30-day thromboembolic event (n=6) and the 30-day mortality (n=6), 2,3% respectively. Three of the six patients died of non-cardiovascular cause. One of the 30-day thromboembolic event resulted in mortality at day 31. Table 2 summarize the data related to the thromboembolic events. There was no statistically significant predictor of any of the two endpoints when comparing the patients with event and patients without. Conclusion This study shows that the risk of thromboembolic events after TOE-guided cardioversion is not negligible, 2,3% within a 30-day follow up. The 30-day mortality was 2,3%, half of which were non-cardiovascular. To analyse potential clinical predictors for any events, larger studies specifically powered for such a question is needed.
Read full abstract